Control method of intelligent watering cart

Abstract

The invention relates to the technical field of watering carts, in particular to a control method of an intelligent watering cart, the watering cart is provided with a laser range finder and a high-definition camera, and an electric control rotating motor and a stop valve are arranged at an elbow of a hedging nozzle, and the control method comprises the steps that S1, road width data are collected in real time through a laser range finder sensor; meanwhile, road image information is obtained through a high-definition camera group; s2, extracting road condition features, pedestrian positions, obstacle positions and dust concentration grades; s3, performing fusion analysis to generate a road environment model; s4, a preset watering strategy library is matched according to the road environment model, a control instruction set is dynamically generated, S5, the control instruction set is issued to an execution mechanism, and the terrain clearance, the flow and the opening and closing state of an opposite flushing nozzle are regulated and controlled in real time. The industrial pain points that a traditional watering cart cannot flush away from a wide road and water overflows from a narrow road are thoroughly solved, waste of water resources is avoided, and the service life of the watering cart is prolonged. And the watering work is more accurate in time.

Description

Technical Field

[0001] The present invention relates to the technical field of sprinkler trucks, and in particular to a control method for an intelligent sprinkler truck. Background Art

[0002] Traditional sprinkler trucks have numerous shortcomings during operation. The watering range and volume cannot be precisely controlled according to actual needs. The watering pattern cannot be flexibly adjusted for roads of varying widths and cleanliness levels, which can easily lead to water waste and poor cleaning results. Summary of the Invention

[0003] The present invention aims to solve the problem that the watering range and watering amount are difficult to be accurately controlled according to actual needs, and the watering mode cannot be flexibly adjusted on sections of roads with different widths and different cleanliness levels, which easily leads to waste of water resources or poor cleaning effect. The present invention provides a control method for an intelligent sprinkler truck, which improves the accuracy and safety of the sprinkler truck operation through intelligent control to meet the needs of urban sanitation work.

[0004] The present invention is implemented through the following technical solution: a control method for an intelligent sprinkler truck, wherein the sprinkler truck is equipped with a laser rangefinder and a high-definition camera, and an electrically controlled rotary motor and a shut-off valve are provided at the elbow of the flush nozzle. The control method includes:

[0005] S1, using laser ranging sensors installed on the front, sides, and rear of the sprinkler truck to collect road width data in real time, while also obtaining road image information through a high-definition camera group;

[0006] S2. Performing visual recognition processing on the road image information to extract road condition characteristics, pedestrian locations, obstacle locations, and dust concentration levels;

[0007] S3, integrating the road width data with the extracted road condition characteristics, pedestrian positions, obstacle positions, and dust concentration levels to generate a road environment model;

[0008] S4. Dynamically generate a control instruction set based on the road environment model matching the preset watering strategy library, including:

[0009] Adjust the height of the flushing nozzle from the ground based on the road width to control the flushing width;

[0010] Adjust the water pump speed based on the dust concentration level to control the water spraying volume per unit time;

[0011] Trigger the closing command of the hedge nozzle in the corresponding area based on the pedestrian / obstacle position information;

[0012] S5. Send the control instruction set to the actuator through the CAN bus. The actuator includes an electric-controlled rotary motor, a pneumatic shut-off valve, and a water pump, which adjust the height above the ground, flow rate, and opening and closing status of the flush nozzle in real time. The electric-controlled rotary motor adjusts the height above the ground of the flush nozzle; the shut-off valve controls the opening and closing of the flush nozzle; and the water pump controls the amount of water sprayed per unit time.

[0013] Furthermore, in step S2, the dust concentration levels on the road are divided into low dust concentration, medium dust concentration, and high dust concentration through image information recognition. When the dust concentration is judged to be low, the water pump maintains the reference speed; when the dust concentration is judged to be medium, the water pump speed is increased to 120%-150% of the reference value; when the dust concentration is judged to be high, the water pump speed is increased to 180%-200% of the reference value.

[0014] Furthermore, the dust concentration level is divided by the image grayscale variance threshold, specifically including:

[0015] When the grayscale variance satisfies σ²≤500, it is judged as low dust concentration;

[0016] When the grayscale variance satisfies 500<σ²≤2000, it is determined to be in the medium dust concentration range;

[0017] When the grayscale variance satisfies σ²>2000, it is judged as high dust concentration.

[0018] Furthermore, the construction of the road environment model in step S3 adopts a multi-source data spatiotemporal registration algorithm, which specifically includes:

[0019] The laser ranging data is stitched into point clouds based on the vehicle coordinate system;

[0020] Perform perspective transformation on the camera image and map it to the point cloud coordinate system;

[0021] The YOLOv5 model is used to identify dynamic targets in the image and map their three-dimensional coordinates to a point cloud model.

[0022] Furthermore, the electrically controlled rotary motor is provided with an angle encoder, and the actual angle is fed back in real time through the angle encoder.

[0023] Furthermore, a safety priority judgment step is added after step S4: when a pedestrian is identified as entering the 3-meter warning zone on the side of the sprinkler truck, other control instructions are immediately overridden, all flush nozzles are forcibly closed, and the sound and light alarm is activated.

[0024] Furthermore, it also includes historical operation data learning optimization steps: recording the actual watering parameters and cleaning effect scores of different road sections; updating the control parameter mapping table in the watering strategy library through the reinforcement learning algorithm.

[0025] The beneficial effects of the present invention are:

[0026] 1. This invention uses multi-source data fusion of laser ranging and visual recognition, combined with a dynamic matching mechanism of dust concentration-water pump speed and nonlinear control of road width-nozzle height, to completely solve the industry pain points of traditional sprinkler trucks, such as "water cannot be completely flushed on wide roads and water overflows on narrow roads". It also avoids the waste of water resources and enables more accurate watering work.

[0027] 2. The present invention is based on dynamic target recognition and real-time monitoring of the 3-meter warning zone, which can quickly close the corresponding nozzle and activate the sound and light alarm when pedestrians approach, preventing pedestrians from getting wet. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 It is a structural diagram of the waterway of a sprinkler truck in the prior art;

[0029] Figure 2 This is a schematic diagram of the structure of the counter-flow nozzle, the electronically controlled rotary motor, and the shut-off valve used in the present invention;

[0030] Figure 3 This is a system block diagram of the intelligent sprinkler truck of the present invention;

[0031] Figure 4 This is a flow chart of the control method of the intelligent sprinkler truck described in the present invention.

[0032] In the picture:

[0033] 1. Counter-flush nozzle; 2. Electronically controlled rotary motor; 3. Shut-off valve. DETAILED DESCRIPTION

[0034] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0035] like Figure 1 As shown in FIG, it is a structural diagram of the waterway of a sprinkler truck in the prior art. There is a flush nozzle 1 on each side, and the waterways are fixed and cannot be adjusted.

[0036] like Figure 2-4 As shown, a control method for an intelligent sprinkler truck is provided. The sprinkler truck is equipped with a laser rangefinder and a high-definition camera. An electric-controlled rotary motor 2 and a shut-off valve 3 are provided at the elbow of the flushing nozzle 1. The control method includes:

[0037] S1. Laser ranging sensors installed on the front, sides, and rear of the sprinkler truck collect road width data in real time, while high-definition camera groups acquire road image information. This builds a multi-dimensional environmental perception system to address the blind spot detection issues of traditional single sensors and improve data completeness.

[0038] S2. Perform visual recognition processing on the road image information to extract road condition characteristics, pedestrian locations, obstacle locations, and dust concentration levels; convert the original image into quantifiable decision parameters, perform digital analysis of road condition elements, and provide input for intelligent decision-making.

[0039] S3. Fusion analysis of road width data with extracted road condition characteristics, pedestrian locations, obstacle locations, and dust concentration levels generates a road environment model. Multi-source data associations are established to eliminate single-sensor errors and generate a high-precision three-dimensional operational map.

[0040] S4. Dynamically generate a control instruction set based on the road environment model matching the preset watering strategy library, including:

[0041] Adjust the height of the flushing nozzle 1 from the ground based on the road width to control the flushing width;

[0042] Adjust the water pump speed based on the dust concentration level to control the water spraying volume per unit time;

[0043] Based on the pedestrian / obstacle position information, the closing instruction of the hedge nozzle 1 in the corresponding area is triggered; the adaptive matching of operation parameters is achieved to solve the contradiction between resource waste and insufficient cleaning caused by the fixed operation mode.

[0044] S5. The control command set is sent to the actuator via the CAN bus. The actuator includes the electronically controlled rotary motor 2, the pneumatic shut-off valve 3, and the water pump. These actuators control the ground clearance, flow rate, and opening / closing status of the flushing nozzle 1 in real time. The electronically controlled rotary motor 2 adjusts the ground clearance of the flushing nozzle 1; the shut-off valve 3 controls the opening and closing of the flushing nozzle 1; and the water pump controls the water flow rate per unit time. This ensures efficient execution of control commands, with response latency kept to <50ms.

[0045] In practical applications, in step S2, image information is used to identify dust concentration levels on the road, categorizing them into low, medium, and high. When the dust concentration is determined to be low, the water pump maintains its base speed; when it is determined to be medium, the pump speed is increased to 120%-150% of the base speed; and when it is determined to be high, the pump speed is increased to 180%-200% of the base speed. This establishes a dynamic matching mechanism between dust concentration and water volume, improving water efficiency for dust suppression by 35% and avoiding over-watering in low-dust areas.

[0046] In practical applications, dust concentration levels are divided by image grayscale variance thresholds, specifically including:

[0047] When the grayscale variance satisfies σ²≤500, it is judged as low dust concentration;

[0048] When the grayscale variance satisfies 500<σ²≤2000, it is determined to be in the medium dust concentration range;

[0049] When the grayscale variance satisfies σ²>2000, it is determined to be a high dust concentration. A low-cost dust detection solution was developed to replace expensive PM2.5 sensors, reducing hardware costs.

[0050] In practical applications, the construction of the road environment model in step S3 adopts a multi-source data spatiotemporal registration algorithm, which specifically includes:

[0051] The laser ranging data is stitched into point clouds based on the vehicle coordinate system;

[0052] Perform perspective transformation on the camera image and map it to the point cloud coordinate system;

[0053] The YOLOv5 model is used to identify dynamic objects in images and map their 3D coordinates to a point cloud model. This achieves millimeter-level spatial positioning, with a pedestrian position detection error of less than 10cm, a five-fold improvement in accuracy compared to monocular vision.

[0054] In practical applications, the electronically controlled rotary motor 2 is provided with an angle encoder, which provides real-time feedback of the actual angle to improve control accuracy.

[0055] In practice, a safety priority determination step is added after step S4: When a pedestrian is detected entering the 3-meter warning zone to the side of the sprinkler truck, other control instructions are immediately overridden, all flushing nozzles 1 are forcibly closed, and the sound and light alarms are activated. This establishes a safety protection priority mechanism. Compared to traditional sprinkler trucks that require manual driver control, the use of visual recognition automatic control is more precise and reduces the probability of pedestrians getting wet.

[0056] In actual applications, it also includes historical operation data learning and optimization steps: recording the actual watering parameters and cleaning effect scores of different road sections; updating the control parameter mapping table in the watering strategy library through reinforcement learning algorithms, and realizing self-evolution of control strategies through continuous feedback and learning.

[0057] In summary, the control method of an intelligent sprinkler truck described in the present invention can improve the accuracy and safety of sprinkler truck operations through intelligent control, and meet the needs of urban sanitation work.

[0058] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the above embodiments are only for the purpose of illustrating the technical concepts and features of the present invention. Their purpose is to enable those familiar with the art to understand the contents of the present invention and implement them. They are not intended to limit the scope of protection of the present invention. Any equivalent changes or modifications made in accordance with the spirit of the present invention should be included in the scope of protection of the present invention.

Claims

1. A control method for an intelligent sprinkler truck, characterized in that: The water sprinkler is equipped with a laser rangefinder and a high-definition camera. An electric-controlled rotary motor (2) and a shut-off valve (3) are provided at the elbow of the flushing nozzle (1). The control method includes: S1, collects road width data in real time through laser ranging sensors installed on the front, sides, and rear of the sprinkler truck, and obtains road image information through a high-definition camera group; S2. Performing visual recognition processing on the road image information to extract road condition characteristics, pedestrian locations, obstacle locations, and dust concentration levels; S3, integrating the road width data with the extracted road condition characteristics, pedestrian positions, obstacle positions, and dust concentration levels to generate a road environment model; S4. Dynamically generate a control instruction set based on the road environment model matching the preset watering strategy library, including: Adjusting the height of the flushing nozzle (1) from the ground based on the road width to control the flushing width; Adjust the water pump speed based on the dust concentration level to control the water spraying volume per unit time; Based on the pedestrian / obstacle position information, a closing instruction of the hedge nozzle (1) in the corresponding area is triggered; S5. Send the control instruction set to the actuator via the CAN bus. The actuator includes an electric-controlled rotary motor (2), a pneumatic shut-off valve (3), and a water pump, which regulates the height above the ground, the flow rate, and the opening and closing state of the flushing nozzle (1) in real time. The electric-controlled rotary motor (2) adjusts the height above the ground of the flushing nozzle (1); the shut-off valve (3) controls the opening and closing of the flushing nozzle (1); and the water pump controls the water spraying volume per unit time.

2. The control method of an intelligent sprinkler truck according to claim 1, characterized in that: In step S2, the dust concentration levels on the road are divided into low dust concentration, medium dust concentration, and high dust concentration through image information recognition. When the dust concentration is judged to be low, the water pump maintains the reference speed; when the dust concentration is judged to be medium, the water pump speed is increased to 120%-150% of the reference value; when the dust concentration is judged to be high, the water pump speed is increased to 180%-200% of the reference value.

3. The control method of an intelligent sprinkler truck according to claim 2, characterized in that: In step S2, the dust concentration level is divided by the image grayscale variance threshold, specifically including: When the grayscale variance satisfies σ²≤500, it is judged as low dust concentration; When the grayscale variance satisfies 500<σ²≤2000, it is determined to be in the medium dust concentration range; When the grayscale variance satisfies σ²>2000, it is judged as high dust concentration.

4. The control method of an intelligent sprinkler truck according to claim 1, characterized in that: The construction of the road environment model in step S3 adopts a multi-source data spatiotemporal registration algorithm, which specifically includes: The laser ranging data is stitched into point clouds based on the vehicle coordinate system; Perform perspective transformation on the camera image and map it to the point cloud coordinate system; The YOLOv5 model is used to identify dynamic targets in the image and map their three-dimensional coordinates to a point cloud model.

5. The control method of an intelligent sprinkler truck according to claim 1, characterized in that: The electrically controlled rotary motor (2) is provided with an angle encoder, and the actual angle is fed back in real time via the angle encoder.

6. The control method of an intelligent sprinkler truck according to claim 1, characterized in that: A safety priority judgment step is added after step S4: when a pedestrian is identified to enter the 3-meter warning zone on the side of the sprinkler truck, other control instructions are immediately overridden, all the flushing nozzles (1) are forcibly closed, and the sound and light alarm is activated.

7. The control method of an intelligent sprinkler truck according to claim 1, characterized in that: It also includes historical operation data learning optimization steps: recording the actual watering parameters and cleaning effect scores of different road sections; and updating the control parameter mapping table in the watering strategy library through reinforcement learning algorithms.

Images